16:00:39 <dhellmann> #startmeeting oslo 16:00:40 <openstack> Meeting started Fri Jul 18 16:00:39 2014 UTC and is due to finish in 60 minutes. The chair is dhellmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:41 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:43 <openstack> The meeting name has been set to 'oslo' 16:00:48 <dhellmann> our agenda: 16:00:48 <dhellmann> #link https://wiki.openstack.org/wiki/Meetings/Oslo#Agenda_for_Next_Meeting 16:00:57 <dhellmann> #topic red flags from liaisons 16:01:05 <bknudson> keystone -- none 16:01:09 <dhellmann> #info we’re working on the problem with py26 that prevents us from using alpha versions 16:01:53 <dhellmann> mordred thinks that's fixed and we just need to wait for new CI images to be built before we can try some of those patches again 16:01:54 <salv-mobile> Neutron all quiet 16:02:01 <dhellmann> that's what I like to hear :-) 16:02:02 <bknudson> oslo.i18n merged into keystone -- https://review.openstack.org/#/c/104400/ 16:02:12 <dhellmann> excellent! 16:02:28 <mordred> dhellmann: actually... found another issue 16:02:37 <markmc> dhellmann, watching https://jenkins03.openstack.org/job/gate-oslo-incubator-python26/382/console 16:02:41 * dhellmann has a sad 16:02:47 <mordred> yah 16:02:49 <markmc> nope, died 16:03:49 <dhellmann> mordred, markmc : is this just a whack-a-mole issue that we can eventually solve, or is there a fundamental problem with doing this? 16:04:41 <mordred> whack a mole 16:04:56 <mordred> today's problem has nothing to do with Python 16:05:00 <markmc> question is whether we should remove the alphas from global-requirements for now 16:05:05 <markmc> take the pressure off 16:05:16 <mordred> something else in the centos image derped 16:05:27 <dhellmann> markmc: pressure? 16:05:29 <mordred> which borked Nova 16:05:49 <markmc> dhellmann, all the auto-proposed requirements changes that are held up 16:05:57 <markmc> dhellmann, aside from the alphas 16:06:32 <dhellmann> markmc: ah. Can we test this case without those change proposals in place, though? 16:07:03 <markmc> dhellmann, hmm, probably not 16:07:24 <dhellmann> markmc: the other changes can be safely proposed by hand, so we shouldn't be holding anyone up 16:07:45 <dhellmann> at least I think that works 16:07:49 <markmc> dhellmann, well, we can manually propose - the requirements check will fail, but we can see if the py26 succeeds 16:07:54 <dhellmann> true 16:08:00 <markmc> ok, I'll propose 16:08:09 <dhellmann> ok, send me the link and I can get it aproved 16:08:12 <markmc> problem is people don't know what's going on 16:08:25 <dhellmann> we should mail the -dev list, I guess 16:08:46 <dhellmann> do you have time to write a summary? 16:08:48 <bknudson> Can you manually propose if it's not in global requirements? I thought the package wouldn't be in the mirror 16:09:01 <dhellmann> bknudson: the mirror is now a full mirror 16:09:20 <dhellmann> infra switched over to bandersnatch fairly recently 16:09:21 <bknudson> ok 16:09:39 <dhellmann> ok, let's keep going 16:09:43 <dhellmann> #topic adoption status 16:09:43 <dhellmann> #link https://etherpad.openstack.org/p/juno-oslo-adoption-status 16:10:09 <dhellmann> obviously we're a bit blocked on the alphas, but the new libraries aren't alphas so that work can continue 16:10:11 <bknudson> oh, this is where I should say i18n merged to keystone 16:10:37 <dhellmann> bknudson: thanks 16:10:47 <GheRivero> and in ironic 16:11:31 <sreshetnyak> sahara started to migration to oslo.db and oslo.i18n 16:11:33 <dhellmann> GheRivero: if you have a patch to list in the etherpad, that would be good 16:11:44 <GheRivero> oslo.db is on the way, waiting for a change for the opportunistic approach for db migration and testing. 16:11:48 <dhellmann> sreshetnyak: great, can you add that info to the etherpad? 16:11:48 <GheRivero> sure. will do 16:11:53 <dhellmann> GheRivero: thanks 16:12:22 <beekneemech> GheRivero: I think the opportunistic db change merged today. 16:12:40 <i159> beekneemech: yep 16:12:52 <Alexei_9871> I have a patch to migrate cinder to oslo.db 16:12:57 <beekneemech> So next oslo.db release it should be ready to go. 16:13:00 <dhellmann> i159: should we cut a release so ironic can use that fix? 16:13:20 <dhellmann> Alexei_9871: is that the one on line 53 of the etherpad, or another one? 16:13:26 <i159> dhellmann: I think, yes 16:13:29 <sreshetnyak> dhellmann, yes 16:13:38 <dhellmann> i159: ok, maybe monday morning (no releases on friday!) 16:13:46 <Alexei_9871> dhellmann: yes 16:13:48 <viktors> dhellmann: sure :) 16:13:57 <i159> dhellmann: but we still have old version in the global-reqs 16:13:59 <dhellmann> Alexei_9871: perfect, just making sure we had all the info correct 16:14:15 <dhellmann> i159: if that is a minimum version the new version will be picked up automatically 16:14:33 <sreshetnyak> dhellmann, can share link to etherpad? 16:14:39 <dhellmann> https://etherpad.openstack.org/p/juno-oslo-adoption-status 16:14:51 <sreshetnyak> dhellmann, thanks 16:15:19 <i159> dhellmann: ok, that makes my life easier) 16:15:22 <markmc> dhellmann, ok it's https://review.openstack.org/108055 16:15:38 <markmc> dhellmann, tried to capture the summary as I know it so people have something to refer to 16:16:14 <dhellmann> markmc: cool, thanks 16:16:57 <dhellmann> liaisons, let us know if you need help with reviews or if the cores in the other projects have questions about the recommended approach on any of these updates 16:17:07 <dhellmann> let's keep going... 16:17:10 <dhellmann> #topic spec status 16:17:10 <dhellmann> merged several more specs this week 16:17:17 <dhellmann> #info greenio executor for oslo.messaging - https://review.openstack.org/#/c/104792/ - merged 16:17:17 <dhellmann> #info scoping for atoms and flows - https://review.openstack.org/#/c/100048/ - merged 16:17:17 <dhellmann> #info redis backend for taskflow - https://review.openstack.org/#/c/105298/ - merged 16:17:23 <dhellmann> #info AMQP 1.0 driver for oslo.messaging - https://review.openstack.org/#/c/96729/ - merged 16:17:34 <dhellmann> The work on the driver has made some progress, but there's a question about whether or not the distros will have the dependencies set up. I don't know if we're going to have to bump this one to Kilo or not, yet. 16:18:00 <dhellmann> markmc: are you aware of any change in status of those packages for debian or ubuntu? 16:18:33 <markmc> dhellmann, nope, there was a thread on infra list this week 16:18:41 <markmc> dhellmann, one sec and I dig up the link 16:19:14 <markmc> #link http://lists.openstack.org/pipermail/openstack-infra/2014-July/thread.html#1553 16:19:20 <markmc> that's the latest as I know it 16:19:30 <markmc> kgiusti is driving it 16:19:47 <dhellmann> ok 16:20:05 <dhellmann> I'll have to talk to ttx about when we might have to cut that off and retarget for kilo 16:20:18 <dhellmann> we should keep pushing ahead with the work no matter what 16:20:24 <markmc> indeed 16:20:39 <dhellmann> I would like to get some more feedback on the policy configuration directory spec: 16:20:43 <dhellmann> #link https://review.openstack.org/#/c/104157/ 16:20:53 <dhellmann> it seems logical to treat policy files like configuration files 16:21:08 <dhellmann> it's not clear what keystone is planning with a policy fetching API (or why they would do that) 16:21:50 <dhellmann> does anyone have any special interest in the policy code? 16:22:55 * dhellmann takes that as a "no" :-) 16:23:03 <bknudson> it would be ayoung 16:23:07 <dhellmann> other than the policy directory spec, I don’t see any that have an obvious chance of making it into juno, but as we agreed earlier there’s no real need to have a formal “freeze” for specs 16:23:42 <dhellmann> bknudson: do you know whether the API for fetching policies is likely to land in keystone this cycle? 16:23:55 <dhellmann> and, I guess, whether other projects are likely to adopt it? 16:24:14 <bknudson> dhellmann: I think it's unlikely... I don't think there's a spec for it proposed. 16:24:30 <dhellmann> ok, I don't think we should hold up this improvement then 16:24:46 <dhellmann> cores, please add it to your review list 16:24:51 <bknudson> keystone has always had policy fetching... but nobody's used it 16:24:55 <salv-mobile> I don't see demand for an api endpoint to manage authz policies in neutron 16:25:27 * dhellmann nods 16:26:03 <dhellmann> #topic oslo.utils graduation status 16:26:09 <dhellmann> #link https://blueprints.launchpad.net/oslo/+spec/graduate-oslo-utils 16:26:16 <dhellmann> dimsum: any updates? 16:27:07 <dhellmann> #link https://review.openstack.org/#/q/project:openstack/oslo.utils+is:open,n,z 16:27:28 <dhellmann> there are a few reviews open, and several look like they would be release blockers 16:28:08 <dhellmann> looks like one was hit by an unrelated gate failure: https://review.openstack.org/#/c/106753/ 16:28:24 <dhellmann> #topic oslo.serialization graduation status 16:28:29 <dhellmann> #link https://blueprints.launchpad.net/oslo/+spec/graduate-oslo-serialization 16:28:34 <dhellmann> beekneemech: this one is you, right? 16:28:59 <beekneemech> dhellmann: Yeah. See earlier comment about being in demo hell again this week. :-( 16:29:16 <dhellmann> beekneemech: ok, no worries 16:29:20 <beekneemech> I don't know that I'm going to make significant progress on that in the near future now either. 16:29:33 <beekneemech> TripleO meetup next week, and vacation the week after. 16:29:46 <dhellmann> beekneemech: ok, let's talk about handing that off, I might be able to pick it up 16:29:52 <dhellmann> unless we have another volunteer? 16:30:30 <dhellmann> let's see if flaper87|afk has some time, too 16:30:36 <dhellmann> #topic oslo.concurrency graduation status 16:30:40 <dhellmann> #link https://blueprints.launchpad.net/oslo/+spec/graduate-oslo-concurrency 16:30:47 <dhellmann> I assume this is the same? :-) 16:30:53 <beekneemech> Yep 16:31:08 <dhellmann> ok, same answer, let's look for some help 16:31:41 <dhellmann> beekneemech: how about sending email to the ML? 16:31:54 <beekneemech> dhellmann: Okay, will do. 16:32:00 <dhellmann> beekneemech: ok, thanks 16:32:06 <dhellmann> #topic release review 16:32:11 <dhellmann> #info oslo.rootwrap 1.3.0.0a1 - https://etherpad.openstack.org/p/oslo-rootwrap-1.3.0 16:32:32 <dhellmann> did I miss anything, or was that the only release this week? 16:33:11 <dhellmann> markmc is working on oslo.messaging 1.3.1 for ceilometer and i159 is going to release a new oslo.db monday 16:33:54 <dhellmann> #topic juno-2 cut-off next week 16:34:40 <dhellmann> I expect we'll have a harder time with projects accepting new libs as we pass j-2, but we'll keep working on releasing them with the expectation that they will be adopted by k-1 16:34:58 <dhellmann> does that make sense to everyone, or should we be more aggressive about adoption? 16:35:20 <beekneemech> I think that's fine. Our goal was never to get everyone on the libs this cycle, just to have them graduated. 16:35:26 <dhellmann> yep 16:35:50 <beekneemech> Would be nice to WIP adoption patches proposed so they can be rebased early in K and ready to go. 16:36:04 <dhellmann> ooo, good idea, I like that 16:36:12 <bknudson> I'm wondering if keystone should switch to module syncs rather than full oslo-incubator syncs for J2... probably should just continue with the full syncs 16:36:48 <viktors> beekneemech: it's can be hart hard to find reviewers sometimes :( 16:36:50 <dhellmann> bknudson: how far out of sync are you? 16:37:04 <bknudson> dhellmann: keystone was just syncd this week 16:37:11 <bknudson> so we're not out of sync at all 16:37:16 <dhellmann> bknudson: ok, great 16:37:59 <beekneemech> viktors: Right, but it should be much easier if we have patches ready right away in K. 16:38:08 <beekneemech> The longer the adoptions take the more resistance there will be. 16:38:27 <dhellmann> we'll need to reiterate that the incubated version of the modules will be going away at the end of K 16:38:40 <dhellmann> removing the RPC code at the end of juno will help with that signaling :-) 16:38:51 <viktors> beekneemech: I'm affraid, that adoption to to oslo.db will came to Nova only in K ( 16:39:08 <beekneemech> viktors: That's perfectly fine. 16:39:11 <dhellmann> viktors: that's not at all surprising, and why I didn't want us to work on nova before any other projects 16:39:36 <dhellmann> nova is being (rightfully) cautious 16:39:54 <dhellmann> #topic open discussion 16:40:01 <dhellmann> does anyone have anything else they need to bring up this week? 16:40:17 <bknudson> are there any libs released other than those on https://etherpad.openstack.org/p/juno-oslo-adoption-status ? 16:40:18 <viktors> dhellmann, rpodolyaka - can we mark https://blueprints.launchpad.net/oslo/+spec/sync-models-with-migrations BP as approved? 16:40:26 <viktors> sorry, implemented 16:40:41 <dhellmann> bknudson: that list is for the newly created libs, but we have been doing releases of some of the other existing libs 16:40:44 <Alexei_9871> https://review.openstack.org/#/c/105796/ 16:40:50 <Alexei_9871> messaging spec 16:40:55 <dhellmann> bknudson: it looks like oslo.utils is probably the next most likely new library 16:40:57 <bknudson> y, new libs 16:41:01 <Alexei_9871> already implemented 16:41:15 <Alexei_9871> could someone please review it? 16:41:23 <bknudson> ok, keystone would be able to use oslo.utils... for importing 16:41:30 <Alexei_9871> cause it seems that markmc doesn't have enough time to review his own project 16:41:35 <dhellmann> Alexei_9871: it looks like there are a few unanswered questions there. 16:41:45 <Alexei_9871> dhellmann: it's more like nits 16:41:54 <Alexei_9871> so would like to have more votes on this 16:42:10 <Alexei_9871> cause questions are related to how we show diagrams 16:42:43 <dhellmann> Alexei_9871: ok, it looks like the relevant experts are on the review list for that one 16:43:54 <dhellmann> Alexei_9871: and I'll put it on my list, too 16:45:47 <dhellmann> bknudson: the only part of importutils we expect to have public is try_import, I think 16:45:54 <mordred> dhellmann: I'm spinning new nodes now - the problem was actually a rackspace specific yum mirror issue 16:45:56 <dhellmann> the other parts will be there, but not part of the public API 16:46:06 <dhellmann> mordred: good news! 16:46:24 <dhellmann> bknudson: so if you're using, for example, import_class, the expectation is that will switch to use stevedore 16:46:49 <bknudson> always pushing stevedore... 16:47:07 <bknudson> http://git.openstack.org/cgit/openstack/keystone/tree/keystone/common/manager.py#n70 16:47:12 <dhellmann> yes, well, is't the oslo lib, no? 16:47:13 <bknudson> keystone uses import_object 16:47:30 <dhellmann> bknudson: right, those functions will still be available through the incubator 16:47:43 <bknudson> wait, maybe you had a review to switch to stevedore... 16:47:53 <dhellmann> bknudson: I did, and I forget what was wrong with it 16:48:08 <dhellmann> bknudson: https://review.openstack.org/89419 16:48:31 <bknudson> y, let's get that moving. 16:49:05 <dhellmann> bknudson: I'd be happy if you wanted to take that patch over and fix it up to meet the keystone devs' requirements 16:49:26 <bknudson> dhellmann: ok, I'll look at it. 16:49:39 <dhellmann> bknudson: ok, ping me if you have questions you think I can help with 16:49:52 <bknudson> looks pretty straightforward 16:50:00 <dhellmann> cool 16:50:12 <dhellmann> I think that's it for this week? 16:51:00 <dhellmann> ok, thanks for coming, everyone! 16:52:01 <dhellmann> #endmeeting